Announcement

Collapse
No announcement yet.

AMD "RadeonSI" Team Fortress 2 Is Now 75% Faster

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • AMD "RadeonSI" Team Fortress 2 Is Now 75% Faster

    Phoronix: AMD "RadeonSI" Team Fortress 2 Is Now 75% Faster

    The RadeonSI Gallium3D driver for AMD HD 7000 series GPUs and newer is now 75% faster for the Source Engine Team Fortress 2 game thanks to a new patch-set by Marek...

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    Thank you Marek!

    Comment


    • #3
      Thank you Marek! And thank you AMD for committing to making Open Source drivers for your GPU.

      The kids are running TF2 on a HD7750 and an older Athlon X2 with OK performance. However I am looking forward to being able to sustitute Catalyst with the OS drivers :-)

      Comment


      • #4
        Thank you Marek!

        Comment


        • #5
          I guess southern islands shares a lot with northern islands? How different are the architectures anyway?

          Comment


          • #6
            Originally posted by Dukenukemx View Post
            I guess southern islands shares a lot with northern islands? How different are the architectures anyway?
            Don't quote me on that but as far as I know, southern islands uses GCN (Graphics Core Next) that is why a new driver is implemented.

            Comment


            • #7
              Originally posted by Dukenukemx View Post
              I guess southern islands shares a lot with northern islands? How different are the architectures anyway?
              Very different.

              Comment


              • #8
                Originally posted by Dukenukemx View Post
                I guess southern islands shares a lot with northern islands? How different are the architectures anyway?
                Extremely different at a low level (VLIW4 vs GCN architecture).

                However, there is still a lot of higher level optimizations that can be shared.

                Comment


                • #9
                  did anyone try this patches?
                  i have no luck with no.10

                  Code:
                  patching file src/gallium/drivers/r600/Makefile.sources
                  patching file src/gallium/drivers/r600/r600_buffer.c
                  Reversed (or previously applied) patch detected!  Skipping patch.
                  1 out of 1 hunk ignored -- saving rejects to file src/gallium/drivers/r600/r600_buffer.c.rej
                  patching file src/gallium/drivers/r600/r600_pipe.c
                  patching file src/gallium/drivers/r600/r600_pipe.h
                  patching file src/gallium/drivers/r600/r600_state_common.c
                  patching file src/gallium/drivers/radeon/r600_buffer.c
                  patching file src/gallium/drivers/radeon/r600_pipe_common.c
                  patching file src/gallium/drivers/radeon/r600_pipe_common.h
                  patching file src/gallium/drivers/radeonsi/r600_buffer.c
                  patching file src/gallium/drivers/radeonsi/r600_resource.c
                  patching file src/gallium/drivers/radeonsi/r600_translate.c
                  patching file src/gallium/drivers/radeonsi/radeonsi_pipe.c
                  patching file src/gallium/drivers/radeonsi/radeonsi_pipe.h

                  Comment


                  • #10
                    Originally posted by frosth View Post
                    did anyone try this patches?
                    i have no luck with no.10

                    Code:
                    patching file src/gallium/drivers/r600/Makefile.sources
                    patching file src/gallium/drivers/r600/r600_buffer.c
                    Reversed (or previously applied) patch detected!  Skipping patch.
                    1 out of 1 hunk ignored -- saving rejects to file src/gallium/drivers/r600/r600_buffer.c.rej
                    patching file src/gallium/drivers/r600/r600_pipe.c
                    patching file src/gallium/drivers/r600/r600_pipe.h
                    patching file src/gallium/drivers/r600/r600_state_common.c
                    patching file src/gallium/drivers/radeon/r600_buffer.c
                    patching file src/gallium/drivers/radeon/r600_pipe_common.c
                    patching file src/gallium/drivers/radeon/r600_pipe_common.h
                    patching file src/gallium/drivers/radeonsi/r600_buffer.c
                    patching file src/gallium/drivers/radeonsi/r600_resource.c
                    patching file src/gallium/drivers/radeonsi/r600_translate.c
                    patching file src/gallium/drivers/radeonsi/radeonsi_pipe.c
                    patching file src/gallium/drivers/radeonsi/radeonsi_pipe.h
                    That looks like it probably patched ok. Apparently one of the changes has already been made by another patch. Did you try compiling it?

                    Comment

                    Working...
                    X